Artificial intelligence predictive analytics in heart failure: results of the pilot phase of a pragmatic randomized clinical trial.

OBJECTIVES: We conducted an implementation planning process during the pilot phase of a pragmatic trial, which tests an intervention guided by artificial intelligence (AI) analytics sourced from noninvasive monitoring data in heart failure patients (LINK-HF2). MATERIALS AND METHODS: A mixed-method analysis was conducted at 2 pilot sites. Interviews were conducted with 12 of 27 enrolled patients and with 13 participating clinicians. iPARIHS constructs were used for interview construction to identify workflow, communication patterns, and clinician's beliefs. Interviews were transcribed and analyzed using inductive coding protocols to identify key themes. Behavioral response data from the AI-generated notifications were collected. RESULTS: Clinicians responded to notifications within 24 hours in 95% of instances, with 26.7% resulting in clinical action. Four implementation themes emerged: (1) High anticipatory expectations for reliable patient communications, reduced patient burden, and less proactive provider monitoring. (2) The AI notifications required a differential and tailored balance of trust and action advice related to role. (3) Clinic experience with other home-based programs influenced utilization. (4) Responding to notifications involved significant effort, including electronic health record (EHR) review, patient contact, and consultation with other clinicians. DISCUSSION: Clinician's use of AI data is a function of beliefs regarding the trustworthiness and usefulness of the data, the degree of autonomy in professional roles, and the cognitive effort involved. CONCLUSION: The implementation planning analysis guided development of strategies that addressed communication technology, patient education, and EHR integration to reduce clinician and patient burden in the subsequent main randomized phase of the trial. Our results provide important insights into the unique implications of implementing AI analytics into clinical workflow.

Significant Facility-Level Variation in Utilization of and Adherence with Secondary Prevention Therapies Among Patients with Premature Atherosclerotic Cardiovascular Disease: Insights from the VITAL (Veterans wIth premaTure AtheroscLerosis) Registry7.

PURPOSE: We investigated facility-level variation in the use and adherence with antiplatelets and statins among patients with premature and extremely premature ASCVD. METHODS: Using the 2014-2015 nationwide Veterans wIth premaTure AtheroscLerosis (VITAL) registry, we assessed patients with premature (age at first ASCVD event: males < 55 years, females < 65 years) and extremely premature ASCVD (< 40 years). We examined frequency and facility-level variation in any statin, high-intensity statin (HIS), antiplatelet use (aspirin, clopidogrel, ticagrelor, prasugrel, and ticlopidine), and statin adherence (proportion of days covered ≥ 0.8) across 130 nationwide VA healthcare facilities. Facility-level variation was computed using median rate ratios (MRR), a measure of likelihood that two random facilities differ in use of statins or antiplatelets and statin adherence. RESULTS: Our analysis included 135,703 and 7716 patients with premature and extremely premature ASCVD, respectively. Across all facilities, the median (IQR) prescription rate of any statin therapy, HIS therapy, and antiplatelets among patients with premature ASCVD was 0.73 (0.70-0.75), 0.36 (0.32-0.41), and 0.77 (0.73-0.81), respectively. MRR (95% CI) for any statin use, HIS use, and antiplatelet use were 1.53 (1.44-1.60), 1.58 (1.49-1.66), and 1.49 (1.42-1.56), respectively, showing 53, 58, and 49% facility-level variation. The median (IQR) facility-level rate of statin adherence was 0.58 (0.55-0.62) and MRR for statin adherence was 1.13 (1.10-1.15), showing 13% facility-level variation. Similar median facility-level rates and variation were observed among patients with extremely premature ASCVD. CONCLUSIONS: There is suboptimal use and significant facility-level variation in the use of statin and antiplatelet therapy among patients with premature and extremely premature ASCVD. Interventions are needed to optimize care and minimize variation among young ASCVD patients.

Impact of Operator Volumes and Experience on Outcomes After Percutaneous Coronary Intervention: Insights From the Veterans Affairs Clinical Assessment, Reporting and Tracking (CART) Program.

BACKGROUND: Recent analyses of the volume-outcome relationship for percutaneous coronary intervention (PCI) have suggested a less robust association than previously reported. It is unknown if novel factors such as lifetime operator experience influence this relationship. OBJECTIVES: To assess the relationship between annual volumes and outcomes for PCI and determine whether lifetime operator experience modulates the association. METHODS: Annual PCI volumes for facilities and operators within the Veterans Affairs Healthcare System and their relationship with 30-day mortality following PCI were described. The influence of operator lifetime experience on the volume-outcome relationship was assessed. Hierarchical logistic regression was used to adjust for patient and procedural factors. RESULTS: 57,608 PCIs performed from 2013 to 2018 by 382 operators and 63 institutions were analyzed. Operator annualized PCI volume averaged 47.6 (standard deviation [SD] 49.1) and site annualized volume averaged 189.2 (SD 105.2). Median operator experience was 9.0 years (interquartile range [IQR] 4.0-15.0). There was no independent relationship between operator annual volume, institutional volume, or operator lifetime experience with 30-day mortality (p > 0.10). However, the interaction between operator volume and lifetime experience was associated with a marginal decrease in mortality (odds ratio [OR] 0.9998, 95% CI 0.9996-0.9999). CONCLUSIONS: There were no significant associations between facility or operator-level procedural volume and 30-day mortality following PCI in a nationally integrated healthcare system. There was a marginal association between the interaction of operator lifetime experience, operator annual volume, and 30-day mortality that is unlikely to be clinically relevant, though does suggest an opportunity to explore novel factors that may influence the volume-outcome relationship.

Department of Veterans Affairs (VA) and U.S. Department of Defense (DoD) guidelines for management of dyslipidemia and cardiovascular disease risk reduction: Putting evidence in context.

Atherosclerotic cardiovascular disease (ASCVD) remains the leading cause of morbidity and mortality in the United States (U.S.) and incurs significant cost to the healthcare system. Management of cholesterol remains central for ASCVD prevention and has been the focus of multiple national guidelines. In this review, we compare the American Heart Association (AHA)/American College of Cardiology (ACC) and the U.S. Department of Veterans Affairs (VA) and U.S. Department of Defense (DoD) Cholesterol guidelines. We review the evidence base that was used to generate recommendations focusing on 4 distinct themes: 1) the threshold of absolute 10-year ASCVD risk to start a clinician-patient discussion for the initiation of statin therapy in primary prevention patients; 2) the utility of coronary artery calcium score to guide clinician-patient risk discussion pertaining to the initiation of statin therapy for primary ASCVD prevention; 3) the use of moderate versus high-intensity statin therapy in patients with established ASCVD; and 4) the utility of ordering lipid panels after initiation or intensification of lipid lowering therapy to document efficacy and monitor adherence to lipid lowering therapy. We discuss why the VA/DoD and AHA/ACC may have reached different conclusions on these key issues.

Sex-Related Disparities in Cardiovascular Health Care Among Patients With Premature Atherosclerotic Cardiovascular Disease.

Importance: There is a paucity of data regarding secondary prevention care disparities in women with premature and extremely premature atherosclerotic cardiovascular disease (ASCVD), defined as an ASCVD event at 55 years or younger and 40 years or younger, respectively. Objective: To evaluate sex-based differences in antiplatelet agents, any statin, high-intensity statin (HIS) therapy, and statin adherence in patients with premature and extremely premature ASCVD. Design, Setting, and Participants: This was a cross-sectional, multicenter, nationwide VA health care system-based study with patients enrolled in the Veterans With Premature Atherosclerosis (VITAL) registry. The study assessed patients who had at least 1 primary care visit in the Veterans Affairs (VA) health care system from October 1, 2014, to September 30, 2015. Participants included 147 600 veteran patients with premature ASCVD, encompassing ischemic heart disease (IHD), ischemic cerebrovascular disease (ICVD), and peripheral arterial disease (PAD). Exposures: Women vs men with premature and extremely premature ASCVD. Main Outcomes and Measures: Antiplatelet use, any statin use, HIS use, and statin adherence (proportion of days covered [PDC] ≥0.8). Results: We identified 10 413 women and 137 187 men with premature ASCVD (age ≤55 years) and 1340 women and 8145 men with extremely premature (age ≤40 years) ASCVD. Among patients with premature and extremely premature ASCVD, women represented 7.1% and 14.1% of those groups, respectively. When compared with men, women with premature ASCVD had a higher proportion of African American patients (36.1% vs 23.8%) and lower proportions of Asian patients (0.5% vs 0.7%) and White patients (56.1% vs. 68.1%). In the extremely premature ASCVD group, women had a comparatively higher proportion of African American patients (36.8% vs 23.2%) and lower proportion of White patients (55.0% vs 67.8%) and Asian patients (1.3% vs 1.5%) than men. Among patients with premature IHD, women received less antiplatelet (adjusted odds ratio [AOR], 0.47, 95% CI, 0.45-0.50), any statin (AOR, 0.62; 95% CI, 0.59-0.66), and HIS (AOR, 0.63; 95% CI, 0.59-0.66) therapy and were less statin adherent (mean [SD] PDC, 0.68 [0.34] vs 0.73 [0.31]; ß coefficient: -0.02; 95% CI, -0.03 to -0.01) compared with men. Similarly, women with premature ICVD and premature PAD received comparatively less antiplatelet agents, any statin, and HIS. Among patients with extremely premature ASCVD, women also received less antiplatelet therapy (AOR, 0.61; 95% CI, 0.53-0.70), any statin therapy (AOR,0.51; 95% CI, 0.44-0.58), and HIS therapy (AOR, 0.45; 95% CI, 0.37-0.54) than men. There were no sex-associated differences in statin adherence among patients with premature ICVD, premature PAD, or extremely premature ASCVD. Conclusions and Relevance: This cross-sectional study revealed that women veterans with premature ASCVD and extremely premature ASCVD receive less optimal secondary prevention cardiovascular care in comparison with men. Women with premature ASCVD, particularly those with IHD, were also less statin adherent. Multidisciplinary and patient-centered interventions are needed to improve these disparities in women.

Clinical and Anatomic Complexity of Patients Undergoing Coronary Intervention With and Without On-Site Surgical Capabilities: Insights From the Veterans Affairs Clinical Assessment, Reporting and Tracking (CART) Program.

BACKGROUND: Professional society consensus statements articulate the clinical and anatomic complexity of patients that may undergo percutaneous coronary intervention (PCI) without on-site cardiothoracic surgery, although compliance with these recommendations has not been assessed. We sought to evaluate the clinical and anatomic complexity of patients undergoing PCI with and without cardiothoracic surgery on-site. METHODS: We identified all patients undergoing PCI in the Veterans Affairs health care system between October 2009 and September 2017. The clinical and anatomic complexity of patients treated at sites with or without cardiothoracic surgery was evaluated with a comparative interrupted time series, and mortality was ascertained in a propensity-matched cohort. RESULTS: We identified 75 564 patients who underwent PCI, with the majority (53 708, 71%) treated at sites with cardiothoracic surgery. The overall clinical complexity was statistically greater for those treated at sites with cardiothoracic surgery (National Cardiovascular Data Registries CathPCI: 18.4) compared with those at sites without (17.8, P<0.001) throughout the study, with similar annual increases in complexity before (2% versus 3%; P=0.107) and after (3% versus 3%; P=0.704) January 2014. The anatomic complexity of patients treated was also statistically greater (Veterans Affairs SYNTAX: 11.0 versus 10.2; P<0.001) and increased at comparable rates (2% versus 1%, P=0.731) before 2014. After publication of the consensus statement, anatomic complexity declined at sites with cardiothoracic surgery (-2%) but increased at sites without on-site surgery (5%, P=0.025) such that it was similar at the end of the study (P=0.622). Referrals for emergent cardiothoracic surgery were rare regardless of treatment venue (61, 0.08%) and the hazard for mortality was similar (hazard ratio, 0.883 [95% CI, 0.662-1.176]) after propensity matching. CONCLUSIONS: There are minor differences in complexity of patients undergoing coronary intervention at sites with and without cardiothoracic surgery. Clinical outcomes are similar regardless of treatment venue, suggesting an opportunity to improve access to complex interventional care without sacrificing quality.

Continuous Wearable Monitoring Analytics Predict Heart Failure Hospitalization: The LINK-HF Multicenter Study.

BACKGROUND: Implantable cardiac sensors have shown promise in reducing rehospitalization for heart failure (HF), but the efficacy of noninvasive approaches has not been determined. The objective of this study was to determine the accuracy of noninvasive remote monitoring in predicting HF rehospitalization. METHODS: The LINK-HF study (Multisensor Non-invasive Remote Monitoring for Prediction of Heart Failure Exacerbation) examined the performance of a personalized analytical platform using continuous data streams to predict rehospitalization after HF admission. Study subjects were monitored for up to 3 months using a disposable multisensor patch placed on the chest that recorded physiological data. Data were uploaded continuously via smartphone to a cloud analytics platform. Machine learning was used to design a prognostic algorithm to detect HF exacerbation. Clinical events were formally adjudicated. RESULTS: One hundred subjects aged 68.4±10.2 years (98% male) were enrolled. After discharge, the analytical platform derived a personalized baseline model of expected physiological values. Differences between baseline model estimated vital signs and actual monitored values were used to trigger a clinical alert. There were 35 unplanned nontrauma hospitalization events, including 24 worsening HF events. The platform was able to detect precursors of hospitalization for HF exacerbation with 76% to 88% sensitivity and 85% specificity. Median time between initial alert and readmission was 6.5 (4.2-13.7) days. CONCLUSIONS: Multivariate physiological telemetry from a wearable sensor can provide accurate early detection of impending rehospitalization with a predictive accuracy comparable to implanted devices. The clinical efficacy and generalizability of this low-cost noninvasive approach to rehospitalization mitigation should be further tested. Registration: URL: https://www.clinicaltrials.gov. Unique Identifier: NCT03037710.

Engaging Cardiology Providers in Quality Measurement.

See Article Segal et al.

Temporal Trends in Coronary Angiography and Percutaneous Coronary Intervention: Insights From the VA Clinical Assessment, Reporting, and Tracking Program.

OBJECTIVES: The aim of this study was to evaluate temporal trends in characteristics and outcomes among patients referred for invasive coronary procedures within a national health care system for veterans. BACKGROUND: Coronary angiography and percutaneous coronary intervention remain instrumental diagnostic and therapeutic interventions for coronary artery disease. METHODS: All coronary angiographic studies and interventions performed in U.S. Department of Veterans Affairs cardiac catheterization laboratories for fiscal years 2009 through 2015 were identified. The demographic characteristics and management of these patients were stratified by time. Clinical outcomes including readmission (30-day) and mortality were assessed across years. RESULTS: From 2009 to 2015, 194,476 coronary angiographic examinations and 85,024 interventions were performed at Veterans Affairs facilities. The median numbers of angiographic studies (p = 0.81) and interventions (p = 0.22) remained constant over time. Patients undergoing these procedures were progressively older, with more comorbidities, as the proportion classified as having high Framingham risk significantly increased among those undergoing angiography (from 20% to 25%; p < 0.001) and intervention (from 24% to 32%; p < 0.001). Similarly, the median National Cardiovascular Data Registry CathPCI risk score increased for diagnostic (from 14 to 15; p = 0.005) and interventional (from 14 to 18; p = 0.002) procedures. Post-procedural medical management was unchanged over time, although there was increasing adoption of transradial access for diagnostic (from 6% to 36%; p < 0.001) and interventional (from 5% to 32%; p < 0.001) procedures. Complications and clinical outcomes also remained constant, with a trend toward a reduction in the adjusted hazard ratio for percutaneous coronary intervention mortality (hazard ratio: 0.983; 95% confidence interval: 0.967 to 1.000). CONCLUSIONS: Veterans undergoing invasive coronary procedures have had increasing medical complexity over time, without attendant increases in mortality among those receiving interventions. As the Department of Veterans Affairs moves toward a mix of integrated and community-based care, it will be important to account for these demographic shifts so that quality can be maintained.

Likelihood of myocardial infarction during stroke rehabilitation preceded by cardiovascular screening and an exercise tolerance test: the Locomotor Experience Applied Post-Stroke (LEAPS) trial.

BACKGROUND: Coronary artery disease is highly prevalent in patients with stroke, but because revascularization does not improve major clinical outcomes in patients with stable coronary artery disease relative to intensive medical therapy, routine evaluation for this disease is not warranted in stroke patients. However, it might be warranted in patients destined to undergo vigorous physical therapy. The Locomotor Experience Applied Post-Stroke study, a randomized controlled trial of 408 participants that tested the relative efficacy of two rehabilitation techniques on functional walking level, provided the opportunity to address this question. AIM: The study aims to test the efficacy of screening for cardiovascular disease and an exercise tolerance test in assuring safety among patients undergoing vigorous rehabilitation for gait impairment. METHODS: All participants were screened for serious cardiovascular and pulmonary conditions. At six-weeks poststroke, they also completed a cardiovascular screening inventory and underwent an exercise tolerance test involving bicycle ergometry. Participants received 36, 90-min sessions of a prescribed physical therapy (three per week), initiated at either two-months or six-months poststroke. RESULTS: Twenty-nine participants were excluded on the basis of the cardiac screening questionnaire, and 15 failed the exercise tolerance test for cardiovascular reasons. No participant experienced a cardiac event during a treatment session. Two participants experienced myocardial infarctions, but continued in the trial. In three additional participants, myocardial infarctions caused or contributed to death. CONCLUSIONS: The combination of a negative cardiac screen and the absence of exercise tolerance test failure appeared to have a high negative predictive value for cardiac events during treatment, despite the likelihood of a high prevalence of coronary artery disease in our population.

Lemierre's disease: a case with bilateral iliopsoas abscesses and a literature review.

Lemierre's disease is characterized by sepsis, often with an oropharyngeal source, secondary septic emboli and internal jugular vein thrombosis (Lancet 1:701-3, 1936. Clin Microbiol Rev 20(4):622-59, 2007). Septic emboli affecting many bodily sites have been reported, including the lungs, joints, bones, and brain. The case report describes an unusual case of Lemierre's disease in a 64 year old gentleman causing profound sepsis, acute kidney injury, bilateral iliopsoas abscesses and a right hand abscess. To our knowledge, this is the first reported case of Lemierre's disease in the context of bilateral psoas abscesses, and highlights the ambiguity surrounding the definition of Lemierre's disease. The clinical literature review highlights the difficulty in definitively diagnosing the condition and offers some suggestions for recognising and refining the diagnostic criterion of Lemierre's.

Close antiplatelet therapy monitoring and adjustment based upon thrombelastography may reduce late-onset bleeding in HeartMate II recipients.

OBJECTIVES: Bleeding is the most common complication of HeartMate II and is partially attributable to platelet dysfunction; however, antiplatelet therapy is arbitrary in most centres. We investigated how antiplatelet therapy adjustment with thrombelastography affects late-onset bleeding. METHODS: Thrombelastography was used to adjust antiplatelet therapy in 57 HeartMate II recipients. Kaplan-Meier survival curves and Cox proportional hazard ratio model were used to identify predictors of late-onset bleeding in univariate and multivariate analysis. Finally, late-onset bleeding rate in our study was compared with the reported rates in other studies in the literature, all of which did not use any test to monitor or adjust antiplatelet therapy. RESULTS: Mean follow-up was 347 days. Eighteen late-onset bleeding events occurred in 12 patients, a late-onset bleeding rate of 12/57 (21%) or 0.21 events/patient-year. The Kaplan-Meier survival curves demonstrated that late-onset bleeding was more common in the destination therapy cohort (P = 0.02), in patients older than 60 years (P = 0.04) and in females (P = 0.01), none of which was significant in multivariate analysis at a significance level of 0.05. To further investigate the higher bleeding rate in elderly patients, thrombelastography parameters were compared between younger and older patients at the age cut-off of 60 years which demonstrated a prothrombotic change the day after device implantation in younger patients that was absent in the elderly. There was also a trend towards higher requirement for antiplatelet therapy in younger patients while on device support, but the difference did not reach statistical significance. The average late-onset or gastrointestinal bleeding rate among seven comparable studies in the literature that did not use any monitoring test to adjust antiplatelet therapy was 0.49 events/patient-year. CONCLUSIONS: Our study implicates that antiplatelet therapy adjustment with thrombelastography may reduce late-onset bleeding rate in HeartMate II recipients. Bleeding was more common in the elderly recipients and analysis of thrombelastography data suggests that a less aggressive antiplatelet therapy regimen could potentially lower bleeding rate in this vulnerable population.

Does exercise tolerance testing at 60 days poststroke predict rehabilitation performance?

OBJECTIVE: To assess the relationship between exercise tolerance test (ETT) performance at 6 weeks poststroke and subsequent performance in a treadmill and overground locomotor training program (LTP). DESIGN: Prospective cohort study. SETTING: Exercise testing laboratory in either a primary care hospital or outpatient clinic. PARTICIPANTS: Community-dwelling individuals (N=469), 54.9±19.0 days poststroke, enrolled in the Locomotor Experience Applied Post-Stroke randomized controlled trial. INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: For participants randomly assigned to LTP, the number of sessions needed to attain the training goal of 20 minutes of treadmill stepping was determined. Regression analyses determined the contribution of ETT performance (cycling duration), age, and 6-minute walk test (6MWT) distance to attainment of the stepping duration goal. RESULTS: Age, 6MWT, and ETT performance individually accounted for 10.74%, 10.82%, and 10.76%, respectively, of the variance in the number of sessions needed to attain 20 minutes of stepping. When age and 6MWT were included in the model, the additional contribution of ETT performance was rendered nonsignificant (P=.150). CONCLUSIONS: To the extent that ETT performance can be viewed as a measure of cardiovascular fitness rather than neurologic impairment, cardiovascular fitness at the time of the ETT did not make a significant unique contribution to the number of sessions needed to achieve 20 minutes of stepping. The 6MWT, which involves less intensive exercise than the ETT and therefore is likely to be predominantly affected by neurologic impairment and muscular condition, appeared to account for as much variance as the ETT.

The use of left ventricular assist devices in the treatment of class IV heart failure patients: what the non-heart failure specialist needs to know.

The New York Heart Association class IV heart failure patient represents a very sick patient population with a poor 6-month survival rate. With appropriate selection and timing, left ventricular assist devices can provide improved outcomes while these patients are supported for bridge to heart transplant or destination therapy. As the number of heart failure patients supported by left ventricular assist devices increases, physicians will need to be familiar with postoperative chronic issues (such as right ventricular failure, infection, hypertension, atrial fibrillation, bleeding, and thromboembolic events) that can affect the function of these devices.

CXCL5 polymorphisms are associated with variable blood pressure in cardiovascular disease-free adults.

OBJECTIVE: Leukocyte count has been associated with blood pressure, hypertension, and hypertensive complications. We hypothesized that polymorphisms in the CXCL5 gene, which encodes the neutrophilic chemokine ENA-78, are associated with blood pressure in cardiovascular disease (CVD)-free adults and that these polymorphisms are functional. METHODS AND RESULTS: A total of 192 community-dwelling participants without CVD or risk equivalents were enrolled. Two CXCL5 polymorphisms (-156 G > C (rs352046) and 398 G > A (rs425535)) were tested for associations with blood pressure. Allele-specific mRNA expression in leukocytes was also measured to determine whether heterozygosity was associated with allelic expression imbalance. In -156 C variant carriers, systolic blood pressure (SBP) was 7 mmHg higher than in -156 G/G wild-type homozygotes (131 ± 17 vs. 124 ± 14 mmHg; P = 0.008). Similarly, diastolic blood pressure (DBP) was 4 mmHg higher in -156 C variant carriers (78 ± 11 vs. 74 ± 11 mmHg; P = 0.013). In multivariate analysis of SBP, age, sex, body mass index, and the -156 G > C polymorphism were identified as significant variables. Age, sex, and the -156 G > C SNP were further associated with DBP, along with white blood cells. Allelic expression imbalance and significantly higher circulating ENA-78 concentrations were noted for variant carriers. CONCLUSION: CXCL5 gene polymorphisms are functional and associated with variable blood pressure in CVD-free individuals. The role of CXCL5 as a hypertension- and CVD-susceptibility gene should be further explored.